Exploring a Novel Substrate‐Borne Vibratory Signal in the Wolf Spider Schizocosa Floridana

Received: 3 April 2020 | Revised: 23 October 2020 | Accepted: 1 November 2020 DOI: 10.1111/eth.13114

RESEARCH ARTICLE

Exploring a novel substrate-borne vibratory signal in the wolf spider Schizocosa floridana

1University of California, Berkeley, CA, USA Abstract 2University of Nebraska – Lincoln, Lincoln, NE, USA Animals communicate using a diversity of signals produced by a wide array of physi- 3University of California, Davis, CA, USA cal structures. Determining how a signal is produced provides key insights into sig-

Correspondence nal evolution. Here, we examine a complex vibratory mating display produced by Malcolm F. Rosenthal, University of male Schizocosa floridana wolf spiders. This display contains three discrete substrate- California, Berkeley, CA, USA. Email: [email protected] borne acoustic components (known as “thumps”, “taps”, and “chirps”), each of which is anecdotally associated with the movement of a different body part (the pedipalps, Funding information Division of Environmental Biology, Grant/ legs, and abdomen respectively). In order to determine the method of production, Award Number: 1940481; Division of we employ a combination of high-speed video/audio recordings and SEM imaging of Integrative Organismal Systems, Grant/ Award Number: 1556153 and 1556421 possible sound-producing structures. Previous work has suggested that the “chirp” component is tonal, a signal trait that would be potentially unique in the genus. We Editor: Marie Elisabeth Herberstein measured signal tonality for all courtship components, as well as for courtship components from sixteen other Schizocosa wolf spiders. Our results suggest that S. floridana produces courtship song using a combination of shared (palpal stridulation and foreleg percussion) and novel (abdominal movement) sound production mechanisms. Of particular interest, the “chirp”, which is produced using a novel abdominal production mechanism, is the only known tonal signal with acoustic properties that are unique within the genus. We argue that the potential evolution of a novel sound production mechanism has opened up a new axis of signaling trait space in this species, with important implications for how this signal is likely to function and evolve.

KEYWORDS communication, spider, vibratory

Hebets & Papaj, 2005; Pfennig & Pfennig, 2012), less work has fo- 1 | INTRODUCTION cused on how the physical mechanisms that produce these signals might drive their evolution (but see Elias et al., 2006). Understanding Animals produce a wide array of signals intended to modify the be- the physical mechanisms of signal production is critical, however, havior of other individuals to their benefit. The forms of these sig- as the nature of these mechanisms can both constrain and direct nals, which may be used to communicate in contexts ranging from signal evolution (e.g., Derryberry et al., 2012; Montealegre-Z, 2009; mate choice to predator avoidance, are famously diverse. Often, Podos, 2001). this variation is extensive even between closely related species. Specifically, whether new signals arise through new production Yet, while the roles of signaler, receiver, and environment in driving mechanisms or through modifications of pre-existing mechanisms is the divergence and evolution of new animal signals have received of particular interest, as this distinction dictates the constraints and considerable attention (e.g., Endler, 1992; Endler & Basolo, 1998; limitations under which these signals may evolve. Many, perhaps most,

Ethology. 2020;00:1–10. wileyonlinelibrary.com/journal/eth © 2020 Wiley-VCH GmbH | 1 2 | ROSENTHAL et al. of the most commonly studied acoustic signals have arisen through the However, they cannot vary significantly in pitch, or perceived fre- modification of existing signal-producing structures (e.g., Ewing, 1989; quency. Tonal signals thus have the potential to vary along three in- Fitch, 2006). For example, the great diversity of birdsong derives from dependent axes (amplitude, temporal patterning, and pitch) whereas modifications in the use and form of a shared mechanism: the syrinx broadband signals vary along only two. Tonal variation can provide (Catchpole & Slater, 1995; Kingsley et al., 2018; Read & Weary, 1992; information to receivers. For example, pitch differences between Searcy & Andersson, 1986). And because the physical structure of individuals commonly reflect differences in body size (e.g., Gingras the avian vocal tract attenuates harmonic overtones, many bird songs et al., 2012; Hauser, 1993). Likewise, variations in pitch within in- share a similar pure-tone musical sound (Nowicki & Marler, 1988). dividuals can indicate signaler quality (Christie et al., 2004). Tonal These similarities are also potentially evidence of constraints, and the signals can also respond differently to changes in the environment. shared physical bases of the syrinx and of its associated neuromuscu- For example, narrow-bandwidth signals may respond differently to lar structures limit the trait space through which many bird songs are noise than other signals (Raboin & Elias, 2019) and may take advan- able to evolve (Podos et al., 2004). Likewise, for complex signals con- tage of unique spectral transmission properties of their environment taining multiple components, variation between signal components (e.g., McNett & Cocroft, 2008). A truly tonal chirp would thus open produced by the same structure might be limited (e.g., Reichert, 2013), up the possibility of S. floridana song varying in an axis that is known constrained by shared neuromuscular architecture (Arnold, 1992). to be important in many non-spider species, and that is not available Alternatively, novel signals, or signal components, produced by differ- to its congeners. ent production mechanisms may be more able to vary independently. In this study, we explore the spectral properties of the three Arthropod systems are well suited to the study of acoustic sig- major sounds produced by S. floridana. In particular, we focus on nal evolution (including both air- and substrate-borne acoustics; measurements of tonality, as pure-tone signals are unknown in this Ewing, 1989; Hill, 2008). Arthropods produce sound using a number of genus. To establish the presumptive acoustic novelty of the chirp, unique structures (Ewing, 1989; Uhl & Elias, 2011; Virant-Doberlet & we also compare its tonality with measurements made on court- Čokl, 2004), and their rigid exoskeleton means that new sound produc- ship signals from 16 of the 24 described North American species tion structures can evolve anywhere on the body (e.g., Jocqué, 2005; in the genus (Stratton, 2005). Additionally, we explore the poten- Virant-Doberlet & Čokl, 2004). Additionally, many arthropods produce tial mechanisms by which these courtship sounds are produced. complex vibratory songs that utilize the synchronous deployment of We use high-speed video recordings and SEM imaging of putative multiple distinct sound production mechanisms (Virant-Doberlet & sound-producing areas to begin assessing whether novelty in signal Čokl, 2004). We suggest that investigating how individuals produce acoustic properties is associated with modifications of pre-existing distinct signal components within a complex song, whether signal signal-producing mechanisms or the evolution of new structures/ components are produced using the same, different, new, or pre-ex- mechanisms. isting production mechanisms, and how members of one species differ from others in their sound production mechanisms are all key to understanding how complex signals evolve and function. 2 | METHODS Here, we investigate the acoustic properties and production mechanisms of the song of a common North American forest floor 2.1 | Animal collection and care arthropod—the wolf spider Schizocosa floridana (Bryant, 1934). Courting individuals in the wolf spider genus Schizocosa generate We collected immature S. floridana at night in Alachua County, substrate-borne songs to attract mates, with each species producing Florida, over two collection trips. Spiders used in courtship tonality a unique song. But though these songs are often distinguished in measurements and for scanning electron microscopy were collected their temporal patterning (see Hebets et al., 2013; Stratton, 2005), in February 2017. Spiders used in high-speed courtship recording the underlying production mechanisms are mostly common to wolf were collected in January 2019. In both cases, we transported spiders spiders in general (e.g., Hallander, 1967; Rovner, 1967). These mech- to UC Berkeley and individually housed them in 6 cm × 6 cm × 8 cm anisms include percussion of the pedipalps and the front pair of legs, clear plastic containers (Amac Plastic Products). The rearing room in and stridulation via specialized structures on the tibio-cymbial joint which they were housed was maintained on a 12 hr:12 hr light/dark of the pedipalp (Rovner, 1975). Schizocosa floridana song includes schedule at an ambient temperature of 25°C. We fed spiders one repeated production of three acoustic elements. Two of these (the body-size-matched cricket twice per week and provided them with “thump” and the “tap”) are broadband and atonal and are hypoth- ad libitum water. esized to be produced using stridulatory and percussive mechanisms common to the genus (Rosenthal & Hebets, 2012; Rundus et al., 2011). The third component, the “chirp”, appears to be pure 2.2 | Courtship tonality and genus-wide comparison tone, and it has not yet been determined what the mechanism of production is. To measure courtship tonality, we recorded the songs of ten mature Broadband acoustic signals can vary in their amplitude, or in tem- males (identified by the unique palpal morphology associated with poral characteristics such as production rate, duration, or rhythm. sperm transfer) who were induced to court on a stretched nylon ROSENTHAL et al. | 3 substrate impregnated with female silk. Pheromones in female silk elicit We also compared the measurements of S. floridana tonality to spontaneous courtship in male Schizocosa on contact (Kaston, 1936; the tonality of courtship components from sixteen other species in Roberts & Uetz, 2005; Rovner, 1968). We recorded these songs using the genus from previously made recordings (Table S1). These six- a scanning laser vibrometer (Polytec PSV-400, Waldbronn, Germany). teen species represent more than two thirds of all North American We situated the laser point on the nylon surface within 1 millimeter of Schizocosa species (Stratton, 2005). Quantification of tonality was the spider's body and recorded the output from the vibrometer as a performed in the same manner as above, but the number of exem- 24-bit WAV file using Audacity (audacity.sourceforge.net). plar components within an individual, and the number of individu- Using these recordings, we separately measured the tonal- als within each species varied as a result of differences in recording ity of each courtship component within the S. floridana song. For techniques, locations, and times (Table S1). Because these record- each male, we extracted three exemplars of each component ings were made in three different laboratories using different laser (thumps, taps, chirps; see also Figure 1 from Rosenthal et al., 2018 vibrometers and potentially different recording substrates, we chose or Figure 1 from Rosenthal & Elias, 2019). We quantified the spectral not to analyze the differences statistically. We present the differ- entropy from a power spectral density of each isolated component ences for qualitative assessment. exemplar in MATLAB (see Chivers et al., 2017; Giannakopoulos & Pikrakis, 2014; Sueur et al., 2008; Supplemental file S1). Lower entropy values indicate more pure tones, and higher values indicate 2.3 | Movement during sound production more “noisy”, broadband signals. Spectral entropy has been assessed for numerous taxa (e.g.; Chivers et al., 2017; da Silva et al., 2000; We recorded synchronized high-speed video and audio of courting Suzuki et al., 2006) including as a measure of spectral purity (Chivers males in order to match acoustic courtship component production et al., 2017). To assess differences in tonality across signal compo- (i.e. sound) with body movements. We recorded high-speed camera nents, we conducted a linear mixed-effects model with component footage with a Photron Fastcam SA3 (Tokyo, Japan) at 2000 frames tonality as the dependent variable and component type (i.e., thump, per second, paired with recording substrate-borne vibrations using a tap, or chirp) as a fixed effect. Given the inclusion of multiple record- scanning laser vibrometer (Polytec PSV-400, Waldbronn, Germany). ings per male, we included individual identity as a random factor in Males were induced to court using the methods described above, this model. We performed post hoc t tests comparing the estimated and laser recordings were made directly off the stretched nylon sub- marginal means obtained from the mixed model using the “em- strate in the same fashion. We digitized the vibrometer signal (National means” R package. Instruments USB-6251; Austin, TX, USA) and synchronized it with the

FIGURE 1 Spectral entropy (tonality) for the three S. floridana courtship components 4 | ROSENTHAL et al. simultaneously recorded high-speed footage using Midas software buffer, dehydrated in a series of ethanol baths, critical point dried, (v.2.0; Xcitex, Inc.). We made combined video/audio recordings of all and stored in a desiccator cabinet overnight. Samples were mounted three courtship components for four individual males, with record- on a stub using carbon tape and imaged with a Hitachi TM-1000 ings made from both the side-on view (both palps and opisthosoma in SEM at the UC Berkeley Electron Microscope Laboratory. focus) and the front view (palps and forelegs in focus) when possible. We imaged the tibio-cymbial joint of the pedipalp, which We replicated observations across multiple individuals to ensure that is known to be the location of stridulatory structures in related all potential angles had been filmed when needed. For selected court- species (Rovner, 1975), and which we observed to be in motion ship videos, we exported individual frames as.tiff files and identified during some phases of courtship (i.e., during bouts of thumping). landmarks on each spider (e.g., tip of foreleg and tip of opisthosoma). Imaging was done on both intact and dissected joints. We imaged The initial xy coordinates are the 0 position. We then tracked the po- pedipalps in two ways. First, pedipalps were dissected away from sitions of the identified landmarks through the course of a particular the body at the coxa and mounted whole. Second, we dissected signaling behavior using Adobe Photoshop. We modified the number the palpal cymbium (which is the modified tarsus, or last segment) of frames we measured based on the speed of observed movements away from the tibia. We mounted both joints for imaging of the (chirps—every 0.005 s, taps—every 0.005 s, thumps—every 0.03 s). dorsal surface of the cymbium and of the ventral surface of the tibia. Our high-speed videos demonstrated vigorous opisthosomal movement during some phases of courtship (i.e., during chirps) and 2.4 | Scanning electron microscopy so we also imaged the posterior surface of the cephalothorax and the anterior surface of the opisthosoma, because these surfaces We used scanning electron microscopy in order to search for poten- appeared the most likely to come into contact during opisthosomal tial morphological structures associated with sound production. We movement and are also known to be the location of sound-produc- dissected adult male S. floridana specimens under a light microscope. ing structures in other spider species (e.g., Habronattus opisthoso- Samples were fixed in 2% glutaraldehyde in 0.1 M sodium cacodylate mal stridulation, Maddison & Stratton, 1988). Vibration-producing buffer, post-fixed in 1% osmium tetroxide in 0.1 M sodium cacodylate structures at that location in Schizocosa, however, would be

FIGURE 2 Spectral entropy (tonality) for courtship components from seventeen Schizocosa species, including S. floridana, ordered from most to least tonal. Numbers at the end of species names indicate different song components. S. floridana components are shaded in gray ROSENTHAL et al. | 5

FIGURE 3 The coordinated body movements and sounds of a thump, which contains two components: (a) a pronounced vertical movement of the opisthosoma, and horizontal flexion of the pedipalps, followed by (b) small opisthosomal movements and continuing flexion of the pedipalps. Also shown are (i.) changes in signal amplitude, (ii.) vertical movement of palps and opisthosoma, and (iii.) horizontal movement of palps and opisthosoma over the duration of the signal. The frequency spectrum of these two thump components is also shown (iv.) unique. These two body parts were separated through severing of with leg taps being the least tonal, and opisthosomal chirps the most the connecting pedicel. All legs were removed from the cephalo- tonal (Figure 1). Opisthosomal chirps are also the most tonal of any thorax prior to positioning for imaging. Schizocosa signal component measured to date (Figure 2). Thumps were the second most tonal of any Schizocosa component, and leg taps were around average tonality for those Schizocosa species for 3 | RESULTS which we had available recordings.

3.1 | Tonality of courtship components 3.2 | Movement during sound production All three courtship components differ significantly from each other in tonality (measured as their degree of spectral entropy; F2,78 = 351.86, Thumps are composed of two distinct acoustic components p < .0001; pairwise tests: chirp–tap, t = −25.659, p < .0001; chirp– (Figure 3, red and blue regions), (a) a high-amplitude pulse, which is thump, t = −6.998, p < .0001, tap–thump, t = 18.661, p < .0001), primarily associated with vertical movement of the opisthosoma and 6 | ROSENTHAL et al. a significant horizontal flexion of the pedipalps, and (b) an underly- upward movements are fast, taking less than a combined tenth of a ing rumble, which is associated with constant horizontal movements second spanning an angle of deflection of around 66.6° (range 61.5° of the pedipalp. Both of these components are atonal, comprising a to 71.6°). Importantly, the opisthosoma is not observed to be shak- broad bandwidth of frequencies (Figure 1), similar to songs of other ing at the frequency of the chirp itself (8–12 Hz vs. 350 Hz), which previously described Schizocosa species. Taps are also broadband, rules out tremulation as a production mechanism. associated with rapid percussion of the forelegs on the substrate. Neither the pedipalps nor the opisthosoma appears to move during leg taps (Figure 4). 3.3 | Sound production morphology Chirps commonly contain one to three repeated pulses. Each pulse is associated with a single dorsoventral movement of the SEM images of the tibio-cymbial joint of the pedipalp revealed the opisthosoma. Neither the legs nor the palps appear to move during presence of stridulatory structures, with a file on the dorsal sur- chirp production. The downward stroke is swift (0.035 s) and face of the cymbium (Figure 6a,b) and a plectrum on the ventral smooth (Figure 5, red region), producing a narrow-bandwidth tone base of the tibia (Figure 6a,b). This structure is similar in placement (~350 Hz) corresponding to a ~8 Hz opisthosomal movement. The and morphology to the palpal stridulatory organs of other lycosids upward stroke is less smooth, with some vibration of the opistho- (Rovner, 1975). Imaging of the posterior surface of the opisthosoma soma (~12 Hz opisthosomal movement). Both the downward and and the anterior surface of the opisthosoma revealed no structures

FIGURE 4 The coordinated body movements and sounds of a tap, which contains one component: (a) the percussion of a single foreleg against the substrate. (i.) Signal amplitude, and the (ii.) vertical and (iii.) horizontal positions of the foreleg across the duration of the tap are also shown, as is (iv.) the frequency spectrum of the tap ROSENTHAL et al. | 7

FIGURE 5 The coordinated body movements and sounds of a chirp, which contain two components: (a) the smooth downward movement of the second body part (opisthosoma), and (b) the slower, shaky upward movement of the opisthosoma. (i.) Signal amplitude, and the (ii.) vertical and (iii.) horizontal positions of the opisthosoma over the duration of the chirp are also shown, as is (iv.) the frequency spectrum of both the first and second component

known to be associated with sound production, which commonly the different components of S. floridana song to vary independently. requires two interacting sclerotized surfaces. The opisthosomal sur- In support of this idea, previous work has found that the structure of face is entirely soft cuticle, with no evidence of sclerotization, and S. floridana song varies significantly across signaling environments, the cephalothorax is also smooth (Figure 7). with the chirps often responding independently of thumps and taps. For example, chirp rate and duration are correlated with thump and tap rate in some, but not all, light environments (Rosenthal 4 | DISCUSSION et al., 2018). Likewise, S. floridana courtship changes across tempera- tures, but chirps change in a pattern opposite to the other courtship Schizocosa floridana produces multicomponent courtship songs components. Specifically, chirp duration and the number of pulses using the coordinated movements of three body parts. The thumps within a chirp decrease with increasing temperature, whereas all are associated with flexion of the pedipalps, which is consistent with other components increase in rate or duration with increasing tem- stridulation, as well as significant opisthosomal movement. We also perature (Rosenthal & Elias, 2019). This finding suggests interesting uncovered stridulatory structures on the pedipalps that are likely the future avenues of research in the study of complex signal function, mechanism responsible for the production of thumps. The taps are which is often concerned with the relationships between multiple associated with vigorous striking of the forelegs on the substrate and signal components. In particular, we suggest that the functional rela- are therefore most likely purely percussive. No other body parts are tionships between components of a complex signal may be driven (or in motion during the production of the taps. Chirp production is asso- constrained) by the structural relationships underlying signal com- ciated only with the rapid movement of the opisthosoma. However, ponent production. it is not clear how this movement is generating the chirp sound. SEM The chirp component of S. floridana male courtship is obvi- imaging turned up no structures on the prosoma or opisthosoma ously of particular interest. Not only is it acoustically unlike any that are clearly related to sound production. Additionally, the chirp is described Schizocosa courtship component, its method of produc- truly pure tone, a signal type that no documented Schizocosa signal tion remains a mystery. Our findings rule out three of the most production mechanism is known to produce. In fact, the chirp is as commonly used methods for sound production in spiders including pure tone as some bird song (Silva et al., 2000), and potentially more the genus Schizocosa: stridulation, percussion, and tremulation (Uhl tonal than the calls of some crickets (e.g., Chivers et al., 2017). & Elias, 2011). Stridulation is produced using specialized cuticular It is noteworthy that Schizocosa floridana males produce sounds structures. Although we found evidence for these on the pedipalps, through a variety of production mechanisms (e.g. stridulation, per- which are associated with thump production, our high-speed re- cussion, novel opisthosomal movement). We suggest that the lack of cordings revealed that the palps are not moved during chirping, and shared production mechanisms across courtship signals may allow no structures or sclerotized tissue of any kind were found on the 8 | ROSENTHAL et al.

(a) (a)

(b) (b)

FIGURE 6 Electron micrographs of (a) the tybio-cymbial joint, FIGURE 7 Electron micrographs of (a) the posterior-most with both plectrum and file visible, and (b) the dissected cymbium, surface of the prosoma and (b) the anterior-most surface of the with magnified view of the file structure opisthosoma. There are no apparent sound-producing structures on either body part opisthosoma. Neither percussion (striking of one body part against another, or against the substrate) nor tremulation (rapid shaking filtered through resonating areas present on the wings (e.g., Bennet- of a body part that produces vibrations) require specific morphol- Clark, 2003). It is not yet clear what this could be in S. floridana. ogy. However, high-speed video reveals no percussive component Possibly the opisthosoma itself, or some structure within the opist- to the opisthosomal movement observed during chirps (i.e., the hosoma, is being excited to resonate at the frequency of the chirp. opisthosoma does not strike the substrate). Additionally, unlike the Alternatively, it is possible that the chirps of S. floridana are produced observed frequency spectrum of chirps, percussive signals are in- via a “stick and slip” mechanism, which involves the frictional rub- herently broadband (Elias & Mason, 2014). Likewise, the chirp is not bing of soft tissues, similar to the sound produced by pulling a bow tremulatory, as the opisthosoma is not oscillating at the frequency across a violin string (Patek, 2001). Supporting this, the opisthosoma of the chirp. of S. floridana appears to be completely unsclerotized soft tissue, as Although the current study is able to rule out several potential is common in wolf spiders, and the high frequency chirp is produced mechanisms for chirp production, it does not definitively point to- via a single, down-up movement of the opisthosoma. Future work ward any other specific mechanism. Because the single down-up will include ultra-high-speed imaging of the opisthosoma-prosoma opisthosomal movement of the chirp produces a vibration at a much joint to look for evidence of body parts resonating or interacting fric- higher frequency, the structures that produce it must involve some tionally. We also intend to compare prosomal and opisthosomal mor- type of frequency multiplier. One example of such a multiplier would phology of S. floridana with related species to identify potentially be a stridulatory file and scraper, with each sweep of the opistho- novel external structures, and to perform microCT scans for poten- soma drawing a scraper across a file with numerous ridges. However, tial internal structures. It is also worth noting that there is signifi- no such structure was observed in this case. Likewise, because the cant opisthosomal movement during thumps, which are also more chirp is pure tone, we might expect the presence of resonant struc- tonal than most Schizocosa courtship components. Future experi- tures. In crickets, for example, stridulatory chirps are amplified and mental work is necessary to tease apart the potentially interacting ROSENTHAL et al. | 9 contributions of palpal stridulation and opisthosomal movement on IOS—1556153 to EAH). We thank Fernando Montealegre-Z for as- this signal component. sistance with calculating spectral entropy, and members of the Elias While the findings of this study suggest that the chirps may offer laboratory for their valuable feedback on early drafts of this manu- a novel axis of signal phenotypic variation (i.e., frequency), it is not script. We would like to acknowledge Timucua and Seminole peoples yet clear whether frequency information is important in S. floridana on whose ancestral lands we collected S. floridana used in this study. communication. One possibility is that it may encode information on signaler size or body condition, as is the case in many non-spi- ORCID der species (e.g., Anurans: Gingras et al., 2012; Insects: Bennet- Malcolm F. Rosenthal https://orcid.org/0000-0003-1291-1728 Clark, 1998; Birds: Ryan & Brenowitz, 1985; Primates: Hauser, 1993). Rowan McGinley https://orcid.org/0000-0003-3593-4754 The production rate of thumps, taps, and chirps are known to affect mate choice (Rosenthal & Elias, 2019; Rosenthal & Hebets, 2012; REFERENCES Rosenthal et al., 2018; Rundus et al., 2011), but do not reflect male Arnold, S. J. (1992). Constraints on phenotypic evolution. American size or condition (Rosenthal & Hebets, 2012). It is possible that S. Naturalist, 140, S85–S107. https://doi.org/10.1086/285398 Bennet-Clark, H. C. (1998). Size and scale effects as constraints in insect floridana body size is encoded in chirp frequency rather than in rate, sound communication. Philosophical Transactions of the Royal Society and we are currently engaged in work testing this idea as well as ex- B, 353, 407–419. https://doi.org/10.1098/rstb.1998.0219 ploring other potential roles of this novel acoustic component. Bennet-Clark, H. C. (2003). Wing resonances in the Australian field cricket Teleogryllus oceanicus. Journal of Experimental Biology, 206, 1479–1496. https://doi.org/10.1242/jeb.00281 Bryant, E. B. (1934). New Lycosidae from Florida. Psyche, 41, 38–41. 5 | CONCLUSIONS https://doi.org/10.1155/1934/40858 Catchpole, C. K., & Slater, P. J. B. (1995). Bird Song. Biological Themes and The investigation of new animal signals is, like the signals themselves, Variations. Cambridge University Press. https://doi.org/10.1017/ multicomponent. How do novel signals evolve, how are they pro- cbo9780511754791.009 Chivers, B. D., Jonsson, T., Soulsbury, C. D., & Montealegre-Z, F. (2017). duced, and how do they function? We suggest that whether novel Structural biomechanics determine spectral purity of bush-cricket signals arise through the modification of an existing signal-producing calls. Biology Letters, 13, https://doi.org/10.1098/rsbl.2017.0573 structure or through the evolution of a new signal-producing struc- Christie, P. J., Mennill, D. J., & Ratcliffe, L. M. (2004). Pitch shifts and ture sets the stage for two different evolutionary trajectories. First, song structure indicate male quality in the dawn chorus of black- capped chickadees. Behavioral Ecology and Sociobiology, 55, 341–348. signal trait space can stretch along existing axes of variation through https://doi.org/10.1007/s00265-003-0711-3 the modification of existing structures. This kind of change may be da Silva, M. I., Piqueira, J. R. C., & Vielliard, J. M. E. (2000). Using Shannon consistent with reinforcement, character displacement, or even sim- entropy on measuring the individual variability in the Rufous-bellied ply directional receiver preferences, all of which select for extreme thrus Turdus rufiventris vocal communication. Journal of Theoretical Biology, 207, 57–64. values along currently existing signaling axes. However, new signals Derryberry, E. P., Seddon, N., Claramunt, S., Tobias, J. A., Baker, A., Alexio, will likely inherit some or all of the constraints of the previously ex- A., & Brumfield, R. T. (2012). Correlated evolution of beak morphol- isting signals, and the ability of these signals to evolve will be like- ogy and song in the neotropical woodcreeper radiation. Evolution, 66, wise constrained. Second, with the evolution of new structures, trait 2784–2797. https://doi.org/10.1111/j.1558-5646.2012.01642.x Elias, D. O., Lee, N., Hebets, E. A., & Mason, A. C. (2006). Seismic signal space can expand to new axes of variation, potentially releasing sig- production in a wolf spider: Parallel versus serial multi-component nal form from previous evolutionary constraints. Thus, new signals signals. Journal of Experimental Biology, 209, 1074–1084. https://doi. derived in this way may be unique, bearing little similarity to the phe- org/10.1242/jeb.02104 notypic characteristics of older signals. This single study does not Elias, D. O., & Mason, A. C. (2014). The role of wave and substrate heterogeneity in vibratory communication: Practical issues in studying have the scope to address such broad evolutionary claims, but we the effect of vibratory environments in communication. In Studying consider it intriguing that the only known tonal signal in Schizocosa vibrational communication (pp. 215–247). Springer. https://doi. wolf spiders appears to be produced by a novel mechanism, rather org/10.1007/978-3-662-43607-3_12 than a modification of a pre-existing one. We suggest that the study Endler, J. A. (1992). Signals, signal conditions, and the direction of of signal evolution needs a broader understanding of how signals evolution. American Naturalist, 139, s125–s153. https://doi. org/10.1086/285308 are produced, whether signals are made through the modification Endler, J. A., & Basolo, A. L. (1998). Sensory ecology, receiver biases of pre-existing structures or the evolution of completely new struc- and sexual selection. Trends in Ecology and Evolution, 13, 415–420. tures, and how mate choice operates on these different categories of https://doi.org/10.1016/s0169-5347(98)01471-2 “novel” signals. By exploring the details of the mechanisms of sound Ewing, A. W. (1989). Arthropod bioacoustics: Neurobiology and behaviour. Cornell University Press. https://doi.org/10.5860/choice.28-0285 production, we can examine sexual selection and signal evolution in Fitch, W. T. (2006). Production of vocalizations in mammals. new and valuable ways. Visual Communication, 3, 145. https://doi.org/10.1016/ b0-08-044854-2/00821-x ACKNOWLEDGEMENTS Giannakopoulos, T., & Pikrakis, A. (2014). Introduction to audio analysis: A MATLAB® approach. Academic Press. https://doi.org/10.1016/ This work was supported by grants from the National Science b978-0-08-099388-1.00010-8 Foundation (IOS—1556421 and DEB-1940481 to DOE; 10 | ROSENTHAL et al.

Gingras, B., Boeckle, M., Herbst, C. T., & Fitch, W. T. (2012). Call acoustic Roberts, J. A., & Uetz, G. W. (2005). Information content of female chem- reflect body size across four clades of anurans. Journal of Zoology, ical signals in the wolf spider, Schizocosa ocreata: Male discrimination 289, 143–150. https://doi.org/10.1111/j.1469-7998.2012.00973.x of reproductive state and receptivity. Animal Behaviour, 70, 217–223. Hallander, H., & Hallander, H. (1967). Courtship display and habitat se- https://doi.org/10.1016/j.anbehav.2004.09.026 lection in the wolf spider Pardosa Chelata (OF Müller). Oikos, 18(1), Rosenthal, M. F., & Elias, D. O. (2019). Nonlinear changes in selection on 145–150. https://doi.org/10.2307/3564644 a mating display across a continuous thermal gradient. Proceedings Hauser, M. D. (1993). The evolution of nonhuman primate vocalizations: of the Royal Society B, 286, 20191450. https://doi.org/10.1098/ Effects of phylogeny, body weight, and social context. American rspb.2019.1450 Naturalist, 142, 528–542. https://doi.org/10.1086/285553 Rosenthal, M. F., & Hebets, E. A. (2012). Resource heterogeneity inter- Hebets, E. A., & Papaj, D. R. (2005). Complex signal function: acts with courtship rate to influence mating success in the wolf spi- Developing a framework of testable hypotheses. Behavioral der Schizocosa floridana. Animal Behaviour, 84, 1341–1346. https:// Ecogy and Sociobiology, 57, 197–214. https://doi.org/10.1007/ doi.org/10.1016/j.anbehav.2012.08.028 s00265-004-0865-7 Rosenthal, M. F., Wilkins, M. R., Shizuka, D., & Hebets, E. A. (2018). Hebets, E. A., Vink, C. J., Sullivan-Beckers, L., & Rosenthal, M. F. (2013). Dynamic changes in display architecture and function across envi- The dominance of seismic signaling and selection for signal com- ronments revealed by a systems approach to animal communication. plexity in Schizocosa multimodal courtship displays. Behavioral Evolution, 72, 1134–1145. https://doi.org/10.1111/evo.13448 Ecology and Sociobiology, 67, 1483–1498. https://doi.org/10.1007/ Rovner, J. S. (1967). Acoustic communication in a Lycosid spider (Lycosa s00265-013-1519-4 rabida Walckenaer). Animal Behaviour, 15, 273–281. https://doi. Hill, P. S. M. (2008). Vibrational communication in animals. Harvard org/10.1016/0003-3472(67)90012-7 University Press. Rovner, J. S. (1968). An analysis of display in the Lycosid spider Lycosa Jocqué, R. (2005). Six stridulating organs on one spider (Araneae, rabida Walcknaer. Animal Behaviour, 16, 358–369. https://doi. Zodariidae): Is this the limit? The Journal of Arachnology, 33, 597–603. org/10.1016/0003-3472(68)90021-3 https://doi.org/10.1636/04-107.1 Rovner, J. S. (1975). Sound production by Nearctic wolf spiders: A sub- Kaston, B. J. (1936). The senses involved in the courtship of some vaga- stratum-coupled stridulatory mechanism. Science, 190, 1309–1310. bond spiders. Brooklyn Entomological Society, 16, 97–167. https://doi.org/10.1126/science.190.4221.1309 Kingsley, E. P., Eliason, C. M., Riede, T., Li, Z., Hiscock, T. W., Farnsworth, Rundus, A. S., Sullivan-Beckers, L., Wilgers, D. J., & Hebets, E. A. (2011). M., Thomson, S. L., Goller, F., Tain, C. J., & Clarke, J. A. (2018). Identity Females are choosier in the dark: Environment-dependent reliance and novelty in the avian syrinx. Proceedings of the National Academy on courtship components and its impact on fitness. Evolution, 65, of Sciences, 115, 10209–10217. https://doi.org/10.1073/pnas.18045 268–282. https://doi.org/10.1111/j.1558-5646.2010.01125.x 86115 Ryan, M. J., & Brenowitz, E. A. (1985). The role of body size, phylogeny, Maddison, W. P., & Stratton, G. E. (1988). Sound production and asso- and ambient noise in the evolution of bird song. American Naturalist, ciated morphology in male jumping spiders of the Habronattus agilis 126, 87–100. https://doi.org/10.1086/284398 species group (Araneae, Salticidae). The Journal of Arachnology, 16, Searcy, W. A., & Andersson, M. (1986). Sexual selection and the evolu- 199–211. tion of song. Annual Review of Ecology and Systematics, 17, 507–533. McNett, G. D., & Cocroft, R. B. (2008). Host shifts favor vibrational signal https://doi.org/10.1146/annurev.es.17.110186.002451 divergence in Enchenopa binotata treehoppers. Behavioral Ecology, 19, Stratton, G. E. (2005). Evolution of ornamentation and courtship be- 650–656. https://doi.org/10.1093/beheco/arn017 havior in Schizocosa: Insights from a phylogeny based on morphol- Montealegre-Z, F. (2009). Scale effects and constraints for ogy (Araneae, Lycosidae). The Journal of Arachnology, 33, 347–376. sound production in katydids (Orthoptera: Tettigoniidae): https://doi.org/10.1636/04-80.1 Correlated evolution between morphology and signal param- Sueur, J., Pavoine, S., Hamerlynck, O., & Duvail, S. (2008). Rapid acous- eters. Journal of Evolutionary Biology, 22, 355–366. https://doi. tic survey for biodiversity appraisal. PLoS One, 3, e4065. https://doi. org/10.1111/j.1420-9101.2008.01652.x org/10.1371/journal.pone.0004065 Nowicki, S., & Marler, P. (1988). How do birds sing? Music Perception, 5, Suzuki, R., Buck, J. R., & Tyack, P. L. (2006). Information entropy of hump- 391–426. https://doi.org/10.2307/40285408 back whale songs. Journal of the Acoustical Society of America, 119, Patek, S. N. (2001). Spiny lobsters stick and slip to make sound. Nature, 1849–1866. https://doi.org/10.1121/1.2161827 411, 153. https://doi.org/10.1038/35075656 Uhl, G., & Elias, D. O. (2011). Communication. In M. E. Herberstein Pfennig, D. W., & Pfennig, K. S. (2012). Evolution’s wedge: Competition and (Ed.), Spider behavior: Flexibility and versatility. Cambridge the origins of diversity (Vol. 12). University of California press. https:// University Press (pp. 127–190). https://doi.org/10.1017/cbo97 doi.org/10.1525/california/9780520274181.001.0001 80511974496.006 Podos, J. (2001). Correlated evolution of morphology and vocal signal Virant-Doberlet, M., & Čokl, A. (2004). Vibrational communication in in- structure in Darwin’s finches. Nature, 409, 185–188. https://doi. sects. Neotropical Entomology, 33, 121–134. https://doi.org/10.1016/ org/10.1038/35051570 b978-0-12-374144-8.00271-x Podos, J., Huber, S. K., & Taft, B. (2004). Bird song: The interface of evolution and mechanism. Annual Review of Ecology, Evolution, and Systematics, 35, 55–87. https://doi.org/10.1146/annurev.ecols ys.35.021103.105719 SUPPORTING INFORMATION Raboin, M., & Elias, D. O. (2019). Anthropogenic noise and the bioacous- Additional supporting information may be found online in the tics of terrestrial invertebrates. Journal of Experimental Biology, 222, Supporting Information section. 12. https://doi.org/10.1242/jeb.178749 Read, A. F., & Weary, D. M. (1992). The evolution of bird song - Comparative analyses. Philosophical Transactions of the Royal Society How to cite this article: Rosenthal MF, Hebets EA, McGinley B, 338, 165–187. https://doi.org/10.1098/rstb.1992.0137 Reichert, M. S. (2013). Patterns of variability are consistent across R, et al. Exploring a novel substrate-borne vibratory signal in signal types in the treefrog Dendropsophus ebraccatus. Biological the wolf spider Schizocosa floridana. Ethology. 2020;00:1–10. Journal of the Linnean Society, 109, 131–145. https://doi. https://doi.org/10.1111/eth.13114 org/10.1111/bij.12028